Existing alarm devices, such as the alarm devices used for production line, usually inform an occurrence of a particular event to operators or workers in the plant by way of flashing light. Such a particular event may be an accident affecting the production (for example, equipment may stop working) or a condition affecting the safety of the production (for example, gas leakage and the like). Although the combination of flashing light and alarm sound does well in alerting, the operators may not easily understand instantaneously what the problem is. The operators have to move to where the human machine interface is for an inspection so as to identify and then solve the problem before the production resumes. Therefore, the conventional way mentioned above is not suitable to quickly solve problems.
The use of speech in the workplace such as plant for broadcasting is more intuitive compared with the combination of flashing light and alarm sound. Speeches allow all of the people in the field understanding instantaneously what the specific problem is going on, making a quick response possible. The existing speech broadcasting requires manufacturers to record a great amount of speech data beforehand, which can be realized manually or by text-to-speech software. After the great amount of speech data is recorded, the data is configured and programmed in a controller such as programmable logic controller (PLC). Certain limitations exist for the above way. The recording process is time consuming and the maintenance cost is relatively high. For instance, when speech data is modified or added, previously recorded speeches need to be checked and analyzed, and consistent parameters such as voice and tone are utilized for the new recording.